This document outlines the results of analysis of the results of a survey sent to people who had attended the adaptation futures conference in 2018. The goal is to assess how research into high end climate change (HECC, ie. Anything to do with warming of greater than 2 degrees) has changed since the same survey was taken two years previously. We also look into how the responses compare between different types of organisations and different locations using various techniques.

0.1 Metadata

There were 198 responses to the AF2018 Survey.

0.1.1 Box Plot of Response Times

Note that due to extreme outliers only response times of over 5 minutes and under 90 minutes are considered here. (37 responses omitted)

The mean response time of these respondents was 19 minutes 6 seconds The minimum response time was 5 minutes 8 seconds The maximum response time was 84 minutes 32 seconds

## NULL

0.2 Likert data for 2018 HECC questions

0.3 A comparison of HECC questions 2016 vs. 2018

The chi-square test finds the probability (p) of the distribution of answers to a question remaining the same in 2018 as it was in 2016. Note: This assumes the data is categorical with no repeat answers, so we assume the people answering the survey are not the same in 2016 as in 2018. This is reasonable in a sample size as large as ours.

The Wilcoxon-Mann-Whitney test tests whether there has been a general shift in opinion rather than a change in the proportions of people falling into categories.

0.3.1 Q17.1: In my work I am considering climate changes of more than 2 degrees C

Neither test finds a significant difference in responses for this question.

## 
##  Pearson's Chi-squared test
## 
## data:  Q17.1_summarised[, -c(1)]
## X-squared = 6.7796, df = 3, p-value = 0.07926
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  value by variable (2016, 2018)
## Z = 1.4845, p-value = 0.1377
## alternative hypothesis: true mu is not equal to 0

0.3.2 Q17.2: I am using high end climate change scenarios for decision-making processes in the context of adaptation/mitigation

Neither test finds a significant difference in responses for this question.

## 
##  Pearson's Chi-squared test
## 
## data:  Q17.2_summarised[, -c(1)]
## X-squared = 2.9519, df = 3, p-value = 0.3991
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  value by variable (2016, 2018)
## Z = -0.69402, p-value = 0.4877
## alternative hypothesis: true mu is not equal to 0

0.3.3 Q18.1: The likelihood of high-end climate change now becoming reality has decreased after the Paris agreement

Both the chi-square and WMW tests show a significant shift to disagreement with this statement, with both strongly disagree and disagree increasing in proportion greatly, with these numbers cominng from the neutral category, as the the agree and strongly agree category are largely unchanged.

## 
##  Pearson's Chi-squared test
## 
## data:  Q18.1_summarised[, -c(1)]
## X-squared = 27.43, df = 4, p-value = 1.627e-05
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  value by variable (2016, 2018)
## Z = 3.7362, p-value = 0.0001868
## alternative hypothesis: true mu is not equal to 0

0.3.4 Q18.2: We should wait with large-scale adaptation activities until we have more certainty about the impacts of high-end climate change

This plot shows that while the proportions of neutral, agree and strongly agree have only decreased slightly, there has a been a significant increase in the strongly disagree proportion and a large reduction in the diagree proportion. It seems people disagree more strongly in 2018 than 2016 and the WMW test shows that this shift is statistically significant.

## Warning in chisq.test(Q18.2_summarised[, -c(1)]): Chi-squared approximation
## may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  Q18.2_summarised[, -c(1)]
## X-squared = 19.142, df = 4, p-value = 0.0007369
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  value by variable (2016, 2018)
## Z = 3.3274, p-value = 0.0008766
## alternative hypothesis: true mu is not equal to 0

0.3.5 Q18.3: We, as a society, have tools in place to deal with impacts from high-end climate change

Neither test finds a significant difference in responses for this question.

## 
##  Pearson's Chi-squared test
## 
## data:  Q18.3_summarised[, -c(1)]
## X-squared = 2.7112, df = 4, p-value = 0.6072
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  value by variable (2016, 2018)
## Z = 1.2131, p-value = 0.2251
## alternative hypothesis: true mu is not equal to 0

0.3.6 Q19.1: It is easy to find information about potential impacts of high-end climate change

We see a significant shift from the agree and neutral categories to the disagree category with the extremities changing very little giving a significant result from the chi-square test, however the WMW test finds no significant evidence of an overall shift in opinion.

## 
##  Pearson's Chi-squared test
## 
## data:  Q19.1_summarised[, -c(1)]
## X-squared = 12.602, df = 4, p-value = 0.0134
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  value by variable (2016, 2018)
## Z = 1.5782, p-value = 0.1145
## alternative hypothesis: true mu is not equal to 0

0.3.7 Q19.2: It is easy to find information about decision making with respect to impacts, adaptation and vulnerability in the context of high-end climate change

We see a significant polarisation between 2016 and 2018 with more extreme responses in 2018, giving a significant result in the chi-square test, however there is no evidence of an opinion shift in the WMW test.

## Warning in chisq.test(Q19.2_summarised[, -c(1)]): Chi-squared approximation
## may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  Q19.2_summarised[, -c(1)]
## X-squared = 9.8775, df = 4, p-value = 0.04254
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  value by variable (2016, 2018)
## Z = 0.20946, p-value = 0.8341
## alternative hypothesis: true mu is not equal to 0

0.3.8 An overview of Statistical Test results comparing 2016/18

The Chi-square test has found a significant difference in questions 18.1 (p = 1.812e-05), 18.2(p = 0.0009099), 19.1(p = 0.01172) and 19.2(p = 0.04834). There is a possible issue in 18.2 where too few (3) people selected “strongly agree” to be sure the approximations made in the test are accurate, though much literature says that as long as the number is greater than 1, the approximation is good enough in a large sample.

The Wilcoxon-Mann-Whitney test has found a significant difference in questions 18.1 (p-value = 0.0001868), 18.2 (p-value = 0.0008766) with responses to both of these questions becoming more negative in 2018 than they were in 2016.

0.4 HECC Questions Faceted by Location

0.5 HECC Questions Faceted by Entity

0.6 Analysis of HECC questions by Entity, Governmental vs. Research Organizations

Respondents belonging to entities “Government”, “Government agency” and “Public policy” are categorised as “Gov Related” and are compared to those belonging to the “Research organization” category. Note: Available responses were “Government agency”, “Research organization”, “consultancy”, “NGO”, “Government”, “Business” and “Public policy”.

Both the chi-square and Wilcoxon-Mann-Whitney tests show no evidence that respondents from Governmental and research organizations answered the HECC questions differently.

##                     Entity
## Q17.1                Gov Related Research
##   All the time                 9       25
##   Not at all                   1       11
##   Occasionally                 8       18
##   We are starting to           5       19
## Warning in chisq.test(table(HECC_Entity_grouped[, c(x + 1, 1)])): Chi-
## squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_Entity_grouped[, c(x + 1, 1)])
## X-squared = 2.5166, df = 3, p-value = 0.4723
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by Entity (Gov Related, Research)
## Z = 0.43656, p-value = 0.6624
## alternative hypothesis: true mu is not equal to 0
## 
##                     Entity
## Q17.2                Gov Related Research
##   All the time                 6       24
##   Not at all                   1       17
##   Occasionally                 9       17
##   We are starting to           6       15
## Warning in chisq.test(table(HECC_Entity_grouped[, c(x + 1, 1)])): Chi-
## squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_Entity_grouped[, c(x + 1, 1)])
## X-squared = 5.5661, df = 3, p-value = 0.1347
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by Entity (Gov Related, Research)
## Z = 0.31882, p-value = 0.7499
## alternative hypothesis: true mu is not equal to 0
## 
##                             Entity
## Q18.1                        Gov Related Research
##   Agree                                5       13
##   Disagree                             7       26
##   Neither agree nor disagree           6       15
##   Strongly Agree                       1        3
##   Strongly disagree                    4       17
## Warning in chisq.test(table(HECC_Entity_grouped[, c(x + 1, 1)])): Chi-
## squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_Entity_grouped[, c(x + 1, 1)])
## X-squared = 0.80888, df = 4, p-value = 0.9373
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by Entity (Gov Related, Research)
## Z = 0.55095, p-value = 0.5817
## alternative hypothesis: true mu is not equal to 0
## 
##                             Entity
## Q18.2                        Gov Related Research
##   Agree                                3        5
##   Disagree                             6       26
##   Neither agree nor disagree           3        5
##   Strongly Agree                       2        3
##   Strongly disagree                    9       35
## Warning in chisq.test(table(HECC_Entity_grouped[, c(x + 1, 1)])): Chi-
## squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_Entity_grouped[, c(x + 1, 1)])
## X-squared = 3.1085, df = 4, p-value = 0.5398
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by Entity (Gov Related, Research)
## Z = 0.84734, p-value = 0.3968
## alternative hypothesis: true mu is not equal to 0
## 
##                             Entity
## Q18.3                        Gov Related Research
##   Agree                                6       20
##   Disagree                             8       19
##   Neither agree nor disagree           4       11
##   Strongly Agree                       3        5
##   Strongly disagree                    2       18
## Warning in chisq.test(table(HECC_Entity_grouped[, c(x + 1, 1)])): Chi-
## squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_Entity_grouped[, c(x + 1, 1)])
## X-squared = 3.4923, df = 4, p-value = 0.4791
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by Entity (Gov Related, Research)
## Z = 0.71361, p-value = 0.4755
## alternative hypothesis: true mu is not equal to 0
## 
##                             Entity
## Q19.1                        Gov Related Research
##   Difficult                            9       21
##   Easy                                 5       19
##   Neither easy nor difficult           5       24
##   Very difficult                       2        2
##   Very Easy                            2        7
## Warning in chisq.test(table(HECC_Entity_grouped[, c(x + 1, 1)])): Chi-
## squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_Entity_grouped[, c(x + 1, 1)])
## X-squared = 2.9518, df = 4, p-value = 0.5659
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by Entity (Gov Related, Research)
## Z = -1.357, p-value = 0.1748
## alternative hypothesis: true mu is not equal to 0
## 
##                             Entity
## Q19.2                        Gov Related Research
##   Difficult                           11       28
##   Easy                                 3       15
##   Neither easy nor difficult           7       15
##   Very difficult                       1       12
##   Very Easy                            1        3
## Warning in chisq.test(table(HECC_Entity_grouped[, c(x + 1, 1)])): Chi-
## squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_Entity_grouped[, c(x + 1, 1)])
## X-squared = 3.5478, df = 4, p-value = 0.4707
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by Entity (Gov Related, Research)
## Z = 0.35625, p-value = 0.7217
## alternative hypothesis: true mu is not equal to 0

0.6.1 Now Combine categories to improve chi-square approximation:

We combine “We are starting to” and “Occasionally” into an “In between” category for Q17.1/2, and combine the 5-point likert scales into 3 in the obvious way. This also shows no statistically significant differences between Government and Research organisations in both the chi-square and Wilcoxon-Mann-Whitney tests.

##               Entity
## Q17.1          Gov Related Research
##   All the time           9       25
##   In between            13       37
##   Not at all             1       11
## Warning in chisq.test(table(HECC_Entity_grouped_combined[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_Entity_grouped_combined[, c(x + 1, 1)])
## X-squared = 1.8403, df = 2, p-value = 0.3985
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by Entity (Gov Related, Research)
## Z = 0.6851, p-value = 0.4933
## alternative hypothesis: true mu is not equal to 0
## 
##               Entity
## Q17.2          Gov Related Research
##   All the time           6       24
##   In between            15       32
##   Not at all             1       17
## Warning in chisq.test(table(HECC_Entity_grouped_combined[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_Entity_grouped_combined[, c(x + 1, 1)])
## X-squared = 5.3276, df = 2, p-value = 0.06968
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by Entity (Gov Related, Research)
## Z = 0.50233, p-value = 0.6154
## alternative hypothesis: true mu is not equal to 0
## 
##                             Entity
## Q18.1                        Gov Related Research
##   Agree                                6       16
##   Disagree                            11       43
##   Neither agree nor disagree           6       15
## Warning in chisq.test(table(HECC_Entity_grouped_combined[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_Entity_grouped_combined[, c(x + 1, 1)])
## X-squared = 0.76169, df = 2, p-value = 0.6833
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by Entity (Gov Related, Research)
## Z = 0.48691, p-value = 0.6263
## alternative hypothesis: true mu is not equal to 0
## 
##                             Entity
## Q18.2                        Gov Related Research
##   Agree                                5        8
##   Disagree                            15       61
##   Neither agree nor disagree           3        5
## Warning in chisq.test(table(HECC_Entity_grouped_combined[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_Entity_grouped_combined[, c(x + 1, 1)])
## X-squared = 3.0681, df = 2, p-value = 0.2157
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by Entity (Gov Related, Research)
## Z = 1.5391, p-value = 0.1238
## alternative hypothesis: true mu is not equal to 0
## 
##                             Entity
## Q18.3                        Gov Related Research
##   Agree                                9       25
##   Disagree                            10       37
##   Neither agree nor disagree           4       11
## Warning in chisq.test(table(HECC_Entity_grouped_combined[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_Entity_grouped_combined[, c(x + 1, 1)])
## X-squared = 0.36371, df = 2, p-value = 0.8337
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by Entity (Gov Related, Research)
## Z = 0.065395, p-value = 0.9479
## alternative hypothesis: true mu is not equal to 0
## 
##                             Entity
## Q19.1                        Gov Related Research
##   Difficult                           11       23
##   Easy                                 7       26
##   Neither easy nor difficult           5       24
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_Entity_grouped_combined[, c(x + 1, 1)])
## X-squared = 2.1699, df = 2, p-value = 0.3379
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by Entity (Gov Related, Research)
## Z = -1.1894, p-value = 0.2343
## alternative hypothesis: true mu is not equal to 0
## 
##                             Entity
## Q19.2                        Gov Related Research
##   Difficult                           12       40
##   Easy                                 4       18
##   Neither easy nor difficult           7       15
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_Entity_grouped_combined[, c(x + 1, 1)])
## X-squared = 1.1711, df = 2, p-value = 0.5568
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by Entity (Gov Related, Research)
## Z = -0.13256, p-value = 0.8945
## alternative hypothesis: true mu is not equal to 0

0.7 Analysis of HECC questions by Location, Africa vs. Other

The chi-square test shows a significant difference in answers to Q17.2(p-value = 0.04651) and Q19.1(p-value = 0.01926), however the Wilcoxon-Mann-Whitney test shows this difference does not constitute a significant change in opinion. We do, however, see a significant shift in opinion for Q17.1 (p-value = 0.0482) in the WMW test.

##                     Location
## Q17.1                Africa Other
##   All the time           23    39
##   Not at all             13    11
##   Occasionally           22    16
##   We are starting to     25    15
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1[, c(x + 1, 1)])
## X-squared = 7.7198, df = 3, p-value = 0.05217
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Other)
## Z = -1.9756, p-value = 0.0482
## alternative hypothesis: true mu is not equal to 0
## 
##                     Location
## Q17.2                Africa Other
##   All the time           21    31
##   Not at all             13    19
##   Occasionally           25    17
##   We are starting to     23    12
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1[, c(x + 1, 1)])
## X-squared = 7.9759, df = 3, p-value = 0.04651
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Other)
## Z = -0.36009, p-value = 0.7188
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.1                        Africa Other
##   Agree                          20     8
##   Disagree                       24    38
##   Neither agree nor disagree     16    15
##   Strongly Agree                  3     4
##   Strongly disagree              16    19
## Warning in chisq.test(table(HECC_location_grouped1[, c(x + 1, 1)])): Chi-
## squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1[, c(x + 1, 1)])
## X-squared = 8.5911, df = 4, p-value = 0.07217
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Other)
## Z = 1.8072, p-value = 0.07074
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.2                        Africa Other
##   Agree                           7     5
##   Disagree                       26    26
##   Neither agree nor disagree      8     5
##   Strongly Agree                  5     1
##   Strongly disagree              34    47
## Warning in chisq.test(table(HECC_location_grouped1[, c(x + 1, 1)])): Chi-
## squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1[, c(x + 1, 1)])
## X-squared = 5.6845, df = 4, p-value = 0.224
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Other)
## Z = 1.9301, p-value = 0.0536
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.3                        Africa Other
##   Agree                          20    23
##   Disagree                       22    22
##   Neither agree nor disagree     21    16
##   Strongly Agree                  7     5
##   Strongly disagree              10    17
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1[, c(x + 1, 1)])
## X-squared = 2.9789, df = 4, p-value = 0.5614
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Other)
## Z = 0.91711, p-value = 0.3591
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q19.1                        Africa Other
##   Difficult                      16    31
##   Easy                           24    17
##   Neither easy nor difficult     28    27
##   Very difficult                  8     1
##   Very Easy                       5     7
## Warning in chisq.test(table(HECC_location_grouped1[, c(x + 1, 1)])): Chi-
## squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1[, c(x + 1, 1)])
## X-squared = 11.756, df = 4, p-value = 0.01926
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Other)
## Z = 0.83158, p-value = 0.4056
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q19.2                        Africa Other
##   Difficult                      25    39
##   Easy                           13    13
##   Neither easy nor difficult     26    20
##   Very difficult                 12     9
##   Very Easy                       5     2
## Warning in chisq.test(table(HECC_location_grouped1[, c(x + 1, 1)])): Chi-
## squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1[, c(x + 1, 1)])
## X-squared = 5.5358, df = 4, p-value = 0.2366
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Other)
## Z = 1.3758, p-value = 0.1689
## alternative hypothesis: true mu is not equal to 0

0.7.1 Q17.1: In my work I am considering climate changes of more than 2 degrees C

We see that due to the much higher proportion of respondents from outside Africa selecting “All the time” we have a significantly higher positive response from outside Africa than from in.

0.7.2 Q17.2: I am using high end climate change scenarios for decision-making processes in the context of adaptation/mitigation

Here we see a significant difference due to respondents from Africa selecting less extreme values than respondents from elsewhere.

0.7.3 Q19.1: It is easy to find information about potential impacts of high-end climate change

Here respondents from Africa responded more positively on the whole, however they also had a higher proportion of “very difficult” responses.

0.8 Analysis of HECC questions by Location, Africa vs. Europe

We see no evidence of a statistically significant difference in responses between Africa and Europe.

##                     Location
## Q17.1                Africa Europe
##   All the time           23     14
##   Not at all             13      4
##   Occasionally           22      7
##   We are starting to     25      6
## Warning in chisq.test(table(HECC_location_Africa_Europe[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_Africa_Europe[, c(x + 1, 1)])
## X-squared = 3.3316, df = 3, p-value = 0.3433
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = -1.4501, p-value = 0.147
## alternative hypothesis: true mu is not equal to 0
## 
##                     Location
## Q17.2                Africa Europe
##   All the time           21     13
##   Not at all             13      8
##   Occasionally           25      4
##   We are starting to     23      5
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_Africa_Europe[, c(x + 1, 1)])
## X-squared = 7.2769, df = 3, p-value = 0.06358
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = -0.50219, p-value = 0.6155
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.1                        Africa Europe
##   Agree                          20      4
##   Disagree                       24     15
##   Neither agree nor disagree     16      4
##   Strongly Agree                  3      1
##   Strongly disagree              16      9
## Warning in chisq.test(table(HECC_location_Africa_Europe[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_Africa_Europe[, c(x + 1, 1)])
## X-squared = 4.8246, df = 4, p-value = 0.3058
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = 1.5201, p-value = 0.1285
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.2                        Africa Europe
##   Agree                           7      1
##   Disagree                       26     11
##   Neither agree nor disagree      8      3
##   Strongly Agree                  5      0
##   Strongly disagree              34     18
## Warning in chisq.test(table(HECC_location_Africa_Europe[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_Africa_Europe[, c(x + 1, 1)])
## X-squared = 3.9035, df = 4, p-value = 0.4192
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = 1.0123, p-value = 0.3114
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.3                        Africa Europe
##   Agree                          20      7
##   Disagree                       22     11
##   Neither agree nor disagree     21      7
##   Strongly Agree                  7      2
##   Strongly disagree              10      5
## Warning in chisq.test(table(HECC_location_Africa_Europe[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_Africa_Europe[, c(x + 1, 1)])
## X-squared = 0.9787, df = 4, p-value = 0.913
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = 0.6682, p-value = 0.504
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q19.1                        Africa Europe
##   Difficult                      16     13
##   Easy                           24      8
##   Neither easy nor difficult     28      8
##   Very difficult                  8      0
##   Very Easy                       5      3
## Warning in chisq.test(table(HECC_location_Africa_Europe[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_Africa_Europe[, c(x + 1, 1)])
## X-squared = 8.2191, df = 4, p-value = 0.08387
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = 0.29131, p-value = 0.7708
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q19.2                        Africa Europe
##   Difficult                      25     14
##   Easy                           13      6
##   Neither easy nor difficult     26     10
##   Very difficult                 12      1
##   Very Easy                       5      1
## Warning in chisq.test(table(HECC_location_Africa_Europe[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_Africa_Europe[, c(x + 1, 1)])
## X-squared = 4.3342, df = 4, p-value = 0.3627
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = 0.090236, p-value = 0.9281
## alternative hypothesis: true mu is not equal to 0

0.9 Analysis of HECC questions by Location, North vs. South

Here respondents from Europe, North America and Asia are grouped into a “North” category, and the rest into the “South” category.

The chi square test shows no significant differences in answers to the HECC questions between Northern/Southern hemisphere regions, however there is a significant difference in question 17.1 given by the WMW test (p-value = 0.02344)

##                     Location
## Q17.1                North South
##   All the time          34    29
##   Not at all             8    16
##   Occasionally          13    26
##   We are starting to    14    28
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2[, c(x + 1, 1)])
## X-squared = 6.9272, df = 3, p-value = 0.07425
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = 2.2662, p-value = 0.02344
## alternative hypothesis: true mu is not equal to 0
## 
##                     Location
## Q17.2                North South
##   All the time          25    29
##   Not at all            16    17
##   Occasionally          15    27
##   We are starting to    11    25
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2[, c(x + 1, 1)])
## X-squared = 3.4989, df = 3, p-value = 0.3209
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = -0.046304, p-value = 0.9631
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.1                        North South
##   Agree                          7    21
##   Disagree                      32    33
##   Neither agree nor disagree    13    19
##   Strongly Agree                 4     3
##   Strongly disagree             16    19
## Warning in chisq.test(table(HECC_location_grouped2[, c(x + 1, 1)])): Chi-
## squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2[, c(x + 1, 1)])
## X-squared = 5.4766, df = 4, p-value = 0.2418
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = -1.2244, p-value = 0.2208
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.2                        North South
##   Agree                          5     7
##   Disagree                      22    32
##   Neither agree nor disagree     5     8
##   Strongly Agree                 0     7
##   Strongly disagree             40    42
## Warning in chisq.test(table(HECC_location_grouped2[, c(x + 1, 1)])): Chi-
## squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2[, c(x + 1, 1)])
## X-squared = 6.6331, df = 4, p-value = 0.1566
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = -1.6424, p-value = 0.1005
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.3                        North South
##   Agree                         19    26
##   Disagree                      18    26
##   Neither agree nor disagree    15    22
##   Strongly Agree                 5     7
##   Strongly disagree             14    15
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2[, c(x + 1, 1)])
## X-squared = 0.50436, df = 4, p-value = 0.9731
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = -0.41172, p-value = 0.6805
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q19.1                        North South
##   Difficult                     26    24
##   Easy                          15    26
##   Neither easy nor difficult    23    32
##   Very difficult                 1     8
##   Very Easy                      6     7
## Warning in chisq.test(table(HECC_location_grouped2[, c(x + 1, 1)])): Chi-
## squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2[, c(x + 1, 1)])
## X-squared = 6.1488, df = 4, p-value = 0.1883
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = -0.4739, p-value = 0.6356
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q19.2                        North South
##   Difficult                     34    31
##   Easy                          10    16
##   Neither easy nor difficult    18    29
##   Very difficult                 7    15
##   Very Easy                      2     6
## Warning in chisq.test(table(HECC_location_grouped2[, c(x + 1, 1)])): Chi-
## squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2[, c(x + 1, 1)])
## X-squared = 5.1051, df = 4, p-value = 0.2767
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = -1.2252, p-value = 0.2205
## alternative hypothesis: true mu is not equal to 0

0.9.1 Q17.1: In my work I am considering climate changes of more than 2 degrees C

We see a signicantly more positive response to Q17.1 from Northern responents.

0.9.2 Now Combine categories to improve chi-square approximation:

We now see a statistically significant (p-value = 0.03132 for chi square, p-value = 0.01587 for WMW) difference between North and South in Q17.1(In my work I am considering climate changes of more than 2 degrees C), showing a higher proportion of people in the Northern Hemisphere are considering HECC in their work “All the time” vs. “In between”.

##               Location
## Q17.1          North South
##   All the time    34    29
##   In between      27    54
##   Not at all       8    16
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2_combined[, c(x + 1, 1)])
## X-squared = 6.9272, df = 2, p-value = 0.03132
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = 2.4119, p-value = 0.01587
## alternative hypothesis: true mu is not equal to 0
## 
##               Location
## Q17.2          North South
##   All the time    25    29
##   In between      26    52
##   Not at all      16    17
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2_combined[, c(x + 1, 1)])
## X-squared = 3.285, df = 2, p-value = 0.1935
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = 0.21037, p-value = 0.8334
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.1                        North South
##   Agree                         11    24
##   Disagree                      48    52
##   Neither agree nor disagree    13    19
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2_combined[, c(x + 1, 1)])
## X-squared = 3.0029, df = 2, p-value = 0.2228
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = -1.6932, p-value = 0.09042
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.2                        North South
##   Agree                          5    14
##   Disagree                      62    74
##   Neither agree nor disagree     5     8
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2_combined[, c(x + 1, 1)])
## X-squared = 2.6396, df = 2, p-value = 0.2672
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = -1.5327, p-value = 0.1254
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.3                        North South
##   Agree                         24    33
##   Disagree                      32    41
##   Neither agree nor disagree    15    22
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2_combined[, c(x + 1, 1)])
## X-squared = 0.11503, df = 2, p-value = 0.9441
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = -0.22265, p-value = 0.8238
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q19.1                        North South
##   Difficult                     27    32
##   Easy                          21    33
##   Neither easy nor difficult    23    32
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2_combined[, c(x + 1, 1)])
## X-squared = 0.55255, df = 2, p-value = 0.7586
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = -0.73924, p-value = 0.4598
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q19.2                        North South
##   Difficult                     41    46
##   Easy                          12    22
##   Neither easy nor difficult    18    29
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2_combined[, c(x + 1, 1)])
## X-squared = 1.8229, df = 2, p-value = 0.402
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = -1.3368, p-value = 0.1813
## alternative hypothesis: true mu is not equal to 0

0.9.3 Q17.1: In my work I am considering climate changes of more than 2 degrees C

0.10 HECC 2016 Questions Faceted by Location

0.11 Analysis of 2016 HECC questions by Location, Europe vs. Other

Note: in tables below, the scales are simply ranked from low (1) to high (4/5)

We find no significant evidence that questions were answered differently in Europe in 2016

## Warning in function_list[[k]](value): NAs introduced by coercion

## Warning in function_list[[k]](value): NAs introduced by coercion

## Warning in function_list[[k]](value): NAs introduced by coercion

## Warning in function_list[[k]](value): NAs introduced by coercion

## Warning in function_list[[k]](value): NAs introduced by coercion
##      Location
## Q17.1 Europe Other
##     1      8     6
##     2     24    15
##     3     16     8
##     4     54    17
## Warning in chisq.test(table(HECC_location_grouped1_16_num[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_num[, c(x + 1, 1)])
## X-squared = 3.6434, df = 3, p-value = 0.3026
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = 1.9008, p-value = 0.05733
## alternative hypothesis: true mu is not equal to 0
## 
##      Location
## Q17.2 Europe Other
##     1     18    14
##     2     23    11
##     3     32    10
##     4     26    12
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_num[, c(x + 1, 1)])
## X-squared = 3.3173, df = 3, p-value = 0.3452
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = 1.1866, p-value = 0.2354
## alternative hypothesis: true mu is not equal to 0
## 
##      Location
## Q18.1 Europe Other
##     1     10     0
##     2     33    14
##     3     43    21
##     4     17     9
##     5      4     5
## Warning in chisq.test(table(HECC_location_grouped1_16_num[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_num[, c(x + 1, 1)])
## X-squared = 7.2547, df = 4, p-value = 0.123
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = -1.9314, p-value = 0.05344
## alternative hypothesis: true mu is not equal to 0
## 
##      Location
## Q18.2 Europe Other
##     1     33     9
##     2     49    25
##     3     13     5
##     4     10     8
##     5      2     1
## Warning in chisq.test(table(HECC_location_grouped1_16_num[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_num[, c(x + 1, 1)])
## X-squared = 3.685, df = 4, p-value = 0.4503
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = -1.5222, p-value = 0.1279
## alternative hypothesis: true mu is not equal to 0
## 
##      Location
## Q18.3 Europe Other
##     1     17     3
##     2     25    14
##     3     22     9
##     4     35    18
##     5      7     3
## Warning in chisq.test(table(HECC_location_grouped1_16_num[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_num[, c(x + 1, 1)])
## X-squared = 3.1192, df = 4, p-value = 0.5381
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = -0.81562, p-value = 0.4147
## alternative hypothesis: true mu is not equal to 0
## 
##      Location
## Q19.1 Europe Other
##     1      6     3
##     2     17     6
##     3     38    24
##     4     35    14
##     5      6     0
## Warning in chisq.test(table(HECC_location_grouped1_16_num[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_num[, c(x + 1, 1)])
## X-squared = 4.7701, df = 4, p-value = 0.3117
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = 0.84077, p-value = 0.4005
## alternative hypothesis: true mu is not equal to 0
## 
##      Location
## Q19.2 Europe Other
##     1      9     5
##     2     38    18
##     3     43    20
##     4     13     3
##     5      2     0
## Warning in chisq.test(table(HECC_location_grouped1_16_num[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_num[, c(x + 1, 1)])
## X-squared = 2.2182, df = 4, p-value = 0.6957
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = 0.99008, p-value = 0.3221
## alternative hypothesis: true mu is not equal to 0

0.11.1 Now Combine categories to improve chi-square approximation:

This still shows no significant difference between Europe and elsewhere in 2016

##               Location
## Q17.1          Europe Other
##   All the time     54    17
##   In between       40    23
##   Not at all        8     6
## Warning in chisq.test(table(HECC_location_grouped1_16_combined[, c(x + 1, :
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_combined[, c(x + 1, 1)])
## X-squared = 3.461, df = 2, p-value = 0.1772
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = 1.8533, p-value = 0.06384
## alternative hypothesis: true mu is not equal to 0
## 
##               Location
## Q17.2          Europe Other
##   All the time     26    12
##   In between       55    21
##   Not at all       18    14
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_combined[, c(x + 1, 1)])
## X-squared = 2.689, df = 2, p-value = 0.2607
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = 0.97879, p-value = 0.3277
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.1                        Europe Other
##   Agree                          21    14
##   Disagree                       43    14
##   Neither agree nor disagree     43    21
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_combined[, c(x + 1, 1)])
## X-squared = 2.4981, df = 2, p-value = 0.2868
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = -1.5752, p-value = 0.1152
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.2                        Europe Other
##   Agree                          12     9
##   Disagree                       82    34
##   Neither agree nor disagree     13     5
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_combined[, c(x + 1, 1)])
## X-squared = 1.6233, df = 2, p-value = 0.4441
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = -0.90448, p-value = 0.3657
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.3                        Europe Other
##   Agree                          42    21
##   Disagree                       42    17
##   Neither agree nor disagree     22     9
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_combined[, c(x + 1, 1)])
## X-squared = 0.34441, df = 2, p-value = 0.8418
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = -0.54443, p-value = 0.5861
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q19.1                        Europe Other
##   Difficult                      23     9
##   Easy                           41    14
##   Neither easy nor difficult     38    24
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_combined[, c(x + 1, 1)])
## X-squared = 2.592, df = 2, p-value = 0.2736
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = 0.63366, p-value = 0.5263
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q19.2                        Europe Other
##   Difficult                      47    23
##   Easy                           15     3
##   Neither easy nor difficult     43    20
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_combined[, c(x + 1, 1)])
## X-squared = 1.8557, df = 2, p-value = 0.3954
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = 0.94301, p-value = 0.3457
## alternative hypothesis: true mu is not equal to 0

0.12 Analysis of 2016 HECC questions combined by Location, North vs. South

The WMW test shows that respondents from the north more strongly disagree in 18.2 (We should wait with large-scale adaptation activities until we have more certainty about the impacts of high-end climate change) but no other questions yield statistically significant differences.

## Warning in function_list[[k]](value): NAs introduced by coercion

## Warning in function_list[[k]](value): NAs introduced by coercion

## Warning in function_list[[k]](value): NAs introduced by coercion

## Warning in function_list[[k]](value): NAs introduced by coercion

## Warning in function_list[[k]](value): NAs introduced by coercion
##      Location
## Q17.1 North South
##     1    12     2
##     2    33     6
##     3    19     5
##     4    63     8
## Warning in chisq.test(table(HECC_location_grouped2_16_num[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2_16_num[, c(x + 1, 1)])
## X-squared = 1.4138, df = 3, p-value = 0.7023
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = 0.67309, p-value = 0.5009
## alternative hypothesis: true mu is not equal to 0
## 
##      Location
## Q17.2 North South
##     1    26     6
##     2    29     5
##     3    39     3
##     4    30     8
## Warning in chisq.test(table(HECC_location_grouped2_16_num[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2_16_num[, c(x + 1, 1)])
## X-squared = 3.4671, df = 3, p-value = 0.325
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = -0.067871, p-value = 0.9459
## alternative hypothesis: true mu is not equal to 0
## 
##      Location
## Q18.1 North South
##     1    10     0
##     2    40     7
##     3    56     8
##     4    23     3
##     5     6     3
## Warning in chisq.test(table(HECC_location_grouped2_16_num[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2_16_num[, c(x + 1, 1)])
## X-squared = 4.8224, df = 4, p-value = 0.306
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = -0.90381, p-value = 0.3661
## alternative hypothesis: true mu is not equal to 0
## 
##      Location
## Q18.2 North South
##     1    40     2
##     2    62    12
##     3    15     3
##     4    13     5
##     5     3     0
## Warning in chisq.test(table(HECC_location_grouped2_16_num[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2_16_num[, c(x + 1, 1)])
## X-squared = 6.6302, df = 4, p-value = 0.1568
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = -2.0636, p-value = 0.03906
## alternative hypothesis: true mu is not equal to 0
## 
##      Location
## Q18.3 North South
##     1    19     1
##     2    30     9
##     3    29     2
##     4    43    10
##     5    10     0
## Warning in chisq.test(table(HECC_location_grouped2_16_num[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2_16_num[, c(x + 1, 1)])
## X-squared = 7.9547, df = 4, p-value = 0.09325
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = 0.018861, p-value = 0.985
## alternative hypothesis: true mu is not equal to 0
## 
##      Location
## Q19.1 North South
##     1     7     2
##     2    21     2
##     3    53     9
##     4    40     9
##     5     6     0
## Warning in chisq.test(table(HECC_location_grouped2_16_num[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2_16_num[, c(x + 1, 1)])
## X-squared = 2.6186, df = 4, p-value = 0.6235
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = -0.25262, p-value = 0.8006
## alternative hypothesis: true mu is not equal to 0
## 
##      Location
## Q19.2 North South
##     1    11     3
##     2    48     8
##     3    52    11
##     4    16     0
##     5     2     0
## Warning in chisq.test(table(HECC_location_grouped2_16_num[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped2_16_num[, c(x + 1, 1)])
## X-squared = 4.0255, df = 4, p-value = 0.4026
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (North, South)
## Z = 1.0576, p-value = 0.2903
## alternative hypothesis: true mu is not equal to 0

0.12.1 Q18.2: We should wait with large-scale adaptation activities until we have more certainty about the impacts of high-end climate change

0.12.2 Now Combine categories to improve chi-square approximation:

This finds no significant differences between North/South responses in 2016

##               Location
## Q17.1          Europe Other
##   All the time     54    17
##   In between       40    23
##   Not at all        8     6
## Warning in chisq.test(table(HECC_location_grouped1_16_combined[, c(x + 1, :
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_combined[, c(x + 1, 1)])
## X-squared = 3.461, df = 2, p-value = 0.1772
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = 1.8533, p-value = 0.06384
## alternative hypothesis: true mu is not equal to 0
## 
##               Location
## Q17.2          Europe Other
##   All the time     26    12
##   In between       55    21
##   Not at all       18    14
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_combined[, c(x + 1, 1)])
## X-squared = 2.689, df = 2, p-value = 0.2607
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = 0.97879, p-value = 0.3277
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.1                        Europe Other
##   Agree                          21    14
##   Disagree                       43    14
##   Neither agree nor disagree     43    21
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_combined[, c(x + 1, 1)])
## X-squared = 2.4981, df = 2, p-value = 0.2868
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = -1.5752, p-value = 0.1152
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.2                        Europe Other
##   Agree                          12     9
##   Disagree                       82    34
##   Neither agree nor disagree     13     5
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_combined[, c(x + 1, 1)])
## X-squared = 1.6233, df = 2, p-value = 0.4441
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = -0.90448, p-value = 0.3657
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.3                        Europe Other
##   Agree                          42    21
##   Disagree                       42    17
##   Neither agree nor disagree     22     9
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_combined[, c(x + 1, 1)])
## X-squared = 0.34441, df = 2, p-value = 0.8418
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = -0.54443, p-value = 0.5861
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q19.1                        Europe Other
##   Difficult                      23     9
##   Easy                           41    14
##   Neither easy nor difficult     38    24
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_combined[, c(x + 1, 1)])
## X-squared = 2.592, df = 2, p-value = 0.2736
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = 0.63366, p-value = 0.5263
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q19.2                        Europe Other
##   Difficult                      47    23
##   Easy                           15     3
##   Neither easy nor difficult     43    20
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_grouped1_16_combined[, c(x + 1, 1)])
## X-squared = 1.8557, df = 2, p-value = 0.3954
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Europe, Other)
## Z = 0.94301, p-value = 0.3457
## alternative hypothesis: true mu is not equal to 0

0.13 2016 and 2018 data combined, Africa vs. Europe

Here we have combined the 2016/18 data and we compare the differences between Europe and Africa. Due to the similarity of the responses to questions 17.1/2 in 2016 and 2018, this is a valid way to obtain a population with a large sample size comparing Africa and Europe for these questions. It could be argued that this is not valid for questions with significant differences in responses, but we include analysis of all questions here.

We see a statistically significant difference in Q17.1 (In my work I am considering climate changes of more than 2 degrees C) in both tests, with Europe considering HECC significantly more frequently than Africa

## Warning in function_list[[k]](value): NAs introduced by coercion

## Warning in function_list[[k]](value): NAs introduced by coercion

## Warning in function_list[[k]](value): NAs introduced by coercion

## Warning in function_list[[k]](value): NAs introduced by coercion

## Warning in function_list[[k]](value): NAs introduced by coercion
##      Location
## Q17.1 Africa Europe
##     1     14     12
##     2     25     31
##     3     27     22
##     4     25     68
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_combined_num[, c(x + 1, 1)])
## X-squared = 13.799, df = 3, p-value = 0.003192
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = -2.9793, p-value = 0.00289
## alternative hypothesis: true mu is not equal to 0
## 
##      Location
## Q17.2 Africa Europe
##     1     16     26
##     2     27     27
##     3     25     37
##     4     22     39
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_combined_num[, c(x + 1, 1)])
## X-squared = 2.5778, df = 3, p-value = 0.4614
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = -0.74922, p-value = 0.4537
## alternative hypothesis: true mu is not equal to 0
## 
##      Location
## Q18.1 Africa Europe
##     1     16     19
##     2     26     48
##     3     19     47
##     4     22     21
##     5      4      5
## Warning in chisq.test(table(HECC_location_combined_num[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_combined_num[, c(x + 1, 1)])
## X-squared = 6.8075, df = 4, p-value = 0.1464
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = 0.40626, p-value = 0.6846
## alternative hypothesis: true mu is not equal to 0
## 
##      Location
## Q18.2 Africa Europe
##     1     35     51
##     2     28     60
##     3      9     16
##     4     11     11
##     5      5      2
## Warning in chisq.test(table(HECC_location_combined_num[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_combined_num[, c(x + 1, 1)])
## X-squared = 6.3283, df = 4, p-value = 0.1759
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = 0.49062, p-value = 0.6237
## alternative hypothesis: true mu is not equal to 0
## 
##      Location
## Q18.3 Africa Europe
##     1     10     22
##     2     26     36
##     3     22     29
##     4     23     42
##     5      7      9
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_combined_num[, c(x + 1, 1)])
## X-squared = 1.909, df = 4, p-value = 0.7525
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = 0.19681, p-value = 0.844
## alternative hypothesis: true mu is not equal to 0
## 
##      Location
## Q19.1 Africa Europe
##     1      9      6
##     2     17     30
##     3     31     46
##     4     27     43
##     5      5      9
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_combined_num[, c(x + 1, 1)])
## X-squared = 2.9575, df = 4, p-value = 0.565
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = -0.68131, p-value = 0.4957
## alternative hypothesis: true mu is not equal to 0
## 
##      Location
## Q19.2 Africa Europe
##     1     12     10
##     2     28     52
##     3     31     53
##     4     13     19
##     5      5      3
## Warning in chisq.test(table(HECC_location_combined_num[, c(x + 1, 1)])):
## Chi-squared approximation may be incorrect
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_combined_num[, c(x + 1, 1)])
## X-squared = 4.7901, df = 4, p-value = 0.3095
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = 0.024136, p-value = 0.9807
## alternative hypothesis: true mu is not equal to 0

0.13.1 17.1: In my work I am considering climate changes of more than 2 degrees C

0.13.2 Now Combine categories to improve chi-square approximation:

We again see that only Q17.1 gives a significant difference

##               Location
## Q17.1          Africa Europe
##   All the time     25     68
##   In between       52     53
##   Not at all       14     12
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_combined_combined[, c(x + 1, 1)])
## X-squared = 12.614, df = 2, p-value = 0.001824
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = -3.4471, p-value = 0.0005667
## alternative hypothesis: true mu is not equal to 0
## 
##               Location
## Q17.2          Africa Europe
##   All the time     22     39
##   In between       52     64
##   Not at all       16     26
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_combined_combined[, c(x + 1, 1)])
## X-squared = 1.4612, df = 2, p-value = 0.4816
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = -0.41102, p-value = 0.6811
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.1                        Africa Europe
##   Agree                          26     26
##   Disagree                       42     67
##   Neither agree nor disagree     19     47
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_combined_combined[, c(x + 1, 1)])
## X-squared = 5.5403, df = 2, p-value = 0.06265
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = 0.76201, p-value = 0.4461
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.2                        Africa Europe
##   Agree                          16     13
##   Disagree                       63    111
##   Neither agree nor disagree      9     16
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_combined_combined[, c(x + 1, 1)])
## X-squared = 3.8525, df = 2, p-value = 0.1457
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = 1.5085, p-value = 0.1314
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q18.3                        Africa Europe
##   Agree                          30     51
##   Disagree                       36     58
##   Neither agree nor disagree     22     29
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_combined_combined[, c(x + 1, 1)])
## X-squared = 0.51755, df = 2, p-value = 0.772
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = -0.1297, p-value = 0.8968
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q19.1                        Africa Europe
##   Difficult                      26     36
##   Easy                           32     52
##   Neither easy nor difficult     31     46
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_combined_combined[, c(x + 1, 1)])
## X-squared = 0.22534, df = 2, p-value = 0.8934
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = -0.47341, p-value = 0.6359
## alternative hypothesis: true mu is not equal to 0
## 
##                             Location
## Q19.2                        Africa Europe
##   Difficult                      40     62
##   Easy                           18     22
##   Neither easy nor difficult     31     53
## 
##  Pearson's Chi-squared test
## 
## data:  table(HECC_location_combined_combined[, c(x + 1, 1)])
## X-squared = 0.74596, df = 2, p-value = 0.6887
## 
## 
##  Asymptotic Wilcoxon-Mann-Whitney Test
## 
## data:  Value by location (Africa, Europe)
## Z = 0.34982, p-value = 0.7265
## alternative hypothesis: true mu is not equal to 0

0.14 Cluster Analysis for HECC Questions

Cluster analysis aims to group data so that data points in each cluster are more similar to points in their cluster than they are to those in other clusters. The algorithm used here uses a Jaccard similarity measure to find the best clustering in a data set.

0.14.1 Cluster Analysis for Question 18

The cluster analysis finds 4 clusters, 3 (clusters 1,2,3) of which with high (>0.75) Clusterwise Jaccard bootstrap, indicating stable clusters. Clusters 2 and 3 are in strong agreement for Q18.1 and Q18.2, showing similarly negative responses, but diverging on Q18.3 with cluster 2 strongly agreeing and cluster 3 strongly disagreeing.

## 
## Attaching package: 'fpc'
## The following object is masked from 'package:xtable':
## 
##     xtable
## boot 1 
## boot 2 
## boot 3 
## boot 4 
## boot 5 
## boot 6 
## boot 7 
## boot 8 
## boot 9 
## boot 10 
## boot 11 
## boot 12 
## boot 13 
## boot 14 
## boot 15 
## boot 16 
## boot 17 
## boot 18 
## boot 19 
## boot 20 
## boot 21 
## boot 22 
## boot 23 
## boot 24 
## boot 25 
## boot 26 
## boot 27 
## boot 28 
## boot 29 
## boot 30 
## boot 31 
## boot 32 
## boot 33 
## boot 34 
## boot 35 
## boot 36 
## boot 37 
## boot 38 
## boot 39 
## boot 40 
## boot 41 
## boot 42 
## boot 43 
## boot 44 
## boot 45 
## boot 46 
## boot 47 
## boot 48 
## boot 49 
## boot 50 
## boot 51 
## boot 52 
## boot 53 
## boot 54 
## boot 55 
## boot 56 
## boot 57 
## boot 58 
## boot 59 
## boot 60 
## boot 61 
## boot 62 
## boot 63 
## boot 64 
## boot 65 
## boot 66 
## boot 67 
## boot 68 
## boot 69 
## boot 70 
## boot 71 
## boot 72 
## boot 73 
## boot 74 
## boot 75 
## boot 76 
## boot 77 
## boot 78 
## boot 79 
## boot 80 
## boot 81 
## boot 82 
## boot 83 
## boot 84 
## boot 85 
## boot 86 
## boot 87 
## boot 88 
## boot 89 
## boot 90 
## boot 91 
## boot 92 
## boot 93 
## boot 94 
## boot 95 
## boot 96 
## boot 97 
## boot 98 
## boot 99 
## boot 100
## * Cluster stability assessment *
## Cluster method:  kmeans 
## Full clustering results are given as parameter result
## of the clusterboot object, which also provides further statistics
## of the resampling results.
## Number of resampling runs:  100 
## 
## Number of clusters found in data:  4 
## 
##  Clusterwise Jaccard bootstrap (omitting multiple points) mean:
## [1] 0.8078009 0.8698806 0.8694783 0.6874373
## dissolved:
## [1]  5  4  2 25
## recovered:
## [1] 72 87 86 47

0.14.2 Question 18 Responses by cluster

Note: the two plots are the same data regrouped differently, one by question, one by cluster.

0.14.3 Analysis of Q18 Clusters

Cluster 1 lean towards agreement with every statement, with few exceptions and contains 25 responses.

Cluster 2 contains 52 responses, disagrees strongly with Q18.1/2 but agrees with Q18.3

Cluster 3, containing 48 responses, almost completely disagrees with all Questions.

Cluster 4 was not a stable cluster (clusterwise Jaccard ~ 0.69) so is not particularly meaningful

0.14.4 Repeat 2016 Analysis (Q17.1, Q18.1/3, Q19.1)

In 2016 we saw two clusters, which diverged on H7 (Q18.3). Again we see that Q18.3 is the most divisive question, with clusters agreeing quite well for Q18.1, though we also see some divergence for Q17.1 and Q19.1 also.

## boot 1 
## boot 2 
## boot 3 
## boot 4 
## boot 5 
## boot 6 
## boot 7 
## boot 8 
## boot 9 
## boot 10 
## boot 11 
## boot 12 
## boot 13 
## boot 14 
## boot 15 
## boot 16 
## boot 17 
## boot 18 
## boot 19 
## boot 20 
## boot 21 
## boot 22 
## boot 23 
## boot 24 
## boot 25 
## boot 26 
## boot 27 
## boot 28 
## boot 29 
## boot 30 
## boot 31 
## boot 32 
## boot 33 
## boot 34 
## boot 35 
## boot 36 
## boot 37 
## boot 38 
## boot 39 
## boot 40 
## boot 41 
## boot 42 
## boot 43 
## boot 44 
## boot 45 
## boot 46 
## boot 47 
## boot 48 
## boot 49 
## boot 50 
## boot 51 
## boot 52 
## boot 53 
## boot 54 
## boot 55 
## boot 56 
## boot 57 
## boot 58 
## boot 59 
## boot 60 
## boot 61 
## boot 62 
## boot 63 
## boot 64 
## boot 65 
## boot 66 
## boot 67 
## boot 68 
## boot 69 
## boot 70 
## boot 71 
## boot 72 
## boot 73 
## boot 74 
## boot 75 
## boot 76 
## boot 77 
## boot 78 
## boot 79 
## boot 80 
## boot 81 
## boot 82 
## boot 83 
## boot 84 
## boot 85 
## boot 86 
## boot 87 
## boot 88 
## boot 89 
## boot 90 
## boot 91 
## boot 92 
## boot 93 
## boot 94 
## boot 95 
## boot 96 
## boot 97 
## boot 98 
## boot 99 
## boot 100
## * Cluster stability assessment *
## Cluster method:  kmeans 
## Full clustering results are given as parameter result
## of the clusterboot object, which also provides further statistics
## of the resampling results.
## Number of resampling runs:  100 
## 
## Number of clusters found in data:  2 
## 
##  Clusterwise Jaccard bootstrap (omitting multiple points) mean:
## [1] 0.8103102 0.7838500
## dissolved:
## [1] 15 18
## recovered:
## [1] 73 64